Dietary Therapy In Epilepsy Treatment (DIET-Trial): A Randomised Non-Inferiority Trial Comparing KD, MAD & LGIT for Drug Resistant Epilepsy (S35.006)

Objective: To determine whether Modified Atkins Diet (MAD), and Low-Glycemic-Index-Therapy (LGIT) diet are non-inferior to Ketogenic Diet (KD) among children with drug resistant epilepsy (DRE) Background: KD has been successfully used to treat children with DRE. However, data assessing efficacy of MAD and LGIT as compared to KD is scarce. Design/Methods: One-hundred-seventy children aged between 1 and 15 years, who had ≥4 seizures/month, had failed to respond to ≥2 anti-epileptic-drugs, and had not been treated previously with KD/MAD/LGIT participated in this trial between Apr-2016 and Aug-2017. Children were randomly assigned to receive KD, MAD or LGIT as an add-on to the ongoing antiepileptics. Primary endpoint was percentage seizure reduction from baseline at 24 weeks. The trial was powered to assess non-inferiority of MAD and LGIT compared with KD with predefined non-inferiority margin of −15%. The trial is registered with ClinicalTrials.gov, number NCT02708030. Results: One-hundred-fifty-eight children completed the trial: KD(n=52), MAD(n=52) and LGIT(n=54). After 24 weeks, in per-protocol analysis, the mean(±SD) percentage seizure reduction was −60.3(±32.8) in KD sub-group; −47.9(±45.9) in MAD sub-group; and −54.7(±40) in LGIT sub-group and the difference was statistically insignificant(p=0.18). Mean difference in seizure reduction between the KD and MAD interventions, in the per protocol analysis, was −12.33 (95%CI= −25.16, 0.50); and between KD and LGIT interventions was −5.66 (95%CI= −17.34, 6.02). Adverse events (AE) were noted in 83/170(48.9%) patients; most commonly vomiting (42.4%), diarrhoea (14.5%), constipation (36.5%) and decreased satiety (15.9%). LGIT intervention had significantly lesser AE than KD and MAD interventions(p=0.036). Serious AE were observed in 12/170 (7.1%) patients: metabolic acidosis (pH Conclusions: MAD and LGIT were not non-inferior to KD. Nonetheless, patients on LGIT demonstrated >50% seizure reduction with a better safety profile. Disclosure: Dr. Sondhi has nothing to disclose. Dr. Agarwala has nothing to disclose. Dr. Chakrabarty has nothing to disclose. Dr. Jauhari has nothing to disclose. Dr. Lodha has nothing to disclose. Dr. Pandey has nothing to disclose. Dr. Toteja has nothing to disclose. Dr. Paul has nothing to disclose. Dr. Gulati has nothing to disclose.